Project Summary This application proposes a new U24 Oncology Co-Clinical Imaging Resource entitled VU-PREDICT (Vanderbilt University-PET imaging Resource to Enhance Delivery of Individualized Cancer Therapeutics). Precision cancer medicine, which seeks to exploit unique cellular, molecular and genetic characteristics of individual tumors to optimize treatment, remains a critically unmet need. Despite advances in biomarker technologies that yield high-quality cellular and genomic data, critical gaps remain to consistently match patients with cancer and ideal therapies. While `predictive' genomic assays based on RNA and DNA are now commonplace, current methods largely ignore tumor phenotypes differentiable by quantitative imaging. The overarching vision for VU-PREDICT is a suite of quantitative imaging tools that facilitate the discovery of novel, predictive imaging-derived gene expression signatures; such signatures can be deployed by the greater oncology community to improve the personalization of cancer treatment. The linchpin of VU-PREDICT will be positron emission tomography (PET) imaging. The sensitive and quantitative nature of PET, coupled with the ability to produce biologically active PET tracers, renders PET uniquely capable of both detecting tumors and profiling their specific features. Complementary to genomic approaches, PET imaging provides a quantitative, functional measure of tumor phenotype, and when coupled biopsy approaches, can provide a significantly greater breadth of biological characterization. VU-PREDICT centers on a parallel co-clinical trial of patients with advanced colorectal cancer (CRC) and human-in-mouse PDX (patient-derived xenograft) models. CRC is a leading cause of cancer-related deaths worldwide. Epidermal growth factor receptor (EGFR) neutralizing monoclonal antibodies (mAbs; cetuximab, panitumumab) are approved for treatment of advanced wild-type (WT) RAS CRC. However, durable responses to anti-EGFR mAbs occur in only 12?17% of patients. VU-PREDICT will capitalize upon a Phase I/II clinical trial opening at Vanderbilt combining a glutaminase (GLS1) inhibitor (CB-839, Calithera), EGFR mAb therapy (panitumumab) and irinotecan in patients with advanced and refractory WT RAS CRC. It is anticipated that combining CB-839 with EGFR mAb therapy will resensitize refractory CRC patients with Gln- avid tumors to EGFR blockade. VU-PREDICT will allow our development of quantitative PET imaging measures within this trial and in related preclinical studies that may identify patients likely to respond to combined GLS1/EGFR inhibition. We have four Specific Aims: (1) Optimize quantitative preclinical PET imaging protocols for Gln metabolism; Implement quantitative 18F-FSPG PET (2), 11C-Acetate PET (3), and dual-tracer 11C-Acetate/18F-FSPG PET (4) to discover predictive, imaging-derived gene expression signatures.